Intersection of Passage Lists when ranges are different

As referenced in http://community.logos.com/forums/p/61700/437355.aspx#437355, when merging two Passage Lists for passages which occur in both PLs they don't include references where one of for a single verse and the other is for a range of verses.

For example, merging (based on intersection) the two PLs below only results in one match - Titus 1:4

Is this expected behaviour or should Titus 1:1, 2 & 3 and Titus 2:13 be returned as well?

Find more posts tagged with

Comments

Dave Hooton

It's a strict match, so you should narrow a range like 1-4, 13-14 down to one verse if possible.

Graham Criddle

https://community.logos.com/discussion/comment/437646#Comment_437646

Thanks Dave

It's a strict match,

That's good to know

so you should narrow a range like 1-4, 13-14 down to one verse if possible.

the problem is - in this case - the PL was created from the results of a search where some of the results spanned multiple verses.

Something which probably wouldn't have happened before the Clause search came along

Graham

MJ. Smith

https://community.logos.com/discussion/comment/437749#Comment_437749

I'd made a bug report on this earlier. I would expect the merge to catch cases where the reference in one list is wholly contained in the reference in the other list as well as full matches.

Dave Hooton

https://community.logos.com/discussion/comment/437749#Comment_437749

the problem is - in this case - the PL was created from the results of a search where some of the results spanned multiple verses.

Something which probably wouldn't have happened before the Clause search came along

Yes but:

1:1-4 primarily covers the occurrence of Jesus in v1, but the clause spans vs.1-4 --> delete and manually Add v1

1:4 covers the occurrence at v4

2:13-14 primarily covers the occurrence in v13 --> delete and manually Add v13

2:14 covers the separate clause starting in v14.

Graham Criddle

https://community.logos.com/discussion/comment/437765#Comment_437765

Hi Dave

Yes but:

1:1-4 primarily covers the occurrence of Jesus in v1, but the clause spans vs.1-4 --> delete and manually Add v1

1:4 covers the occurrence at v4

2:13-14 primarily covers the occurrence in v13 --> delete and manually Add v13

2:14 covers the separate clause starting in v14.

Agreed - and it would fix the problem.

However, where this came up (thread referenced in my original post) the original PLs were created from searches across the entire New Testament and included thousands of references (I just used the Titus one to show a problem in a clear format) and I was looking for a way of automatically searching for intersections without having to check for multiple-verse references.

Actually, I hadn't really considered that searches would return multi-verse references in the Clause search so that was interesting to find out[:)]

Graham

Graham Criddle

https://community.logos.com/discussion/comment/437764#Comment_437764

I'd made a bug report on this earlier. I would expect the merge to catch cases where the reference in one list is wholly contained in the reference in the other list as well as full matches.

Thanks MJ[Y]

Dave Hooton

https://community.logos.com/discussion/comment/437771#Comment_437771

I'd made a bug report on this earlier. I would expect the merge to catch cases where the reference in one list is wholly contained in the reference in the other list as well as full matches.

Thanks MJ

You are asking that 1:1-9 be included in the intersection because the other list includes 1:8? With the difference would you be asking that 1:1-9 be removed because the other list includes 1:8? Would 1:8 be removed because the other list has 1:1-9?

MJ. Smith

https://community.logos.com/discussion/comment/437778#Comment_437778

You are asking that 1:1-9 be included in the intersection because the other list includes 1:8? With the difference would you be asking that 1:1-9 be removed because the other list includes 1:8? Would 1:8 be removed because the other list has 1:1-9?

The actual case as I recall, that I reported as a bug one list had 2 & 3 as separate entries, the other 2-3 as a single entry[H].

Bearing in mind that I'm as prejudice against verses and chapters as George is re: interlinears ....If I were designing within the Logos context, I would convert everything to a verse-by-verse comparison although I'd really prefer to use clauses or sentences. However, given the current base design and your example, I would include 1:8 in the intersection because it is within 1:1-9; I would not include 1:1-9 in the intersection because it is not completely contained in 1:8. Similarly I would exclude 1:8 from the difference but include 1:1-7, 9.

Graham Criddle

https://community.logos.com/discussion/comment/437778#Comment_437778

You are asking that 1:1-9 be included in the intersection because the other list includes 1:8? With the difference would you be asking that 1:1-9 be removed because the other list includes 1:8? Would 1:8 be removed because the other list has 1:1-9?

Fair question and I think this is a case where I would like an option allowing me to choose whether I wanted inexact matches for ranges included or not.

I know this would be a UI change but could be useful for these types of cases.

fgh

https://community.logos.com/discussion/comment/437764#Comment_437764

I would expect the merge to catch cases where the reference in one list is wholly contained in the reference in the other list as well as full matches.

So would I!

Dave Hooton

https://community.logos.com/discussion/comment/437782#Comment_437782

However, given the current base design and your example, I would include 1:8 in the intersection because it is within 1:1-9; I would not include 1:1-9 in the intersection because it is not completely contained in 1:8. Similarly I would exclude 1:8 from the difference but include 1:1-7, 9.

It is doable, but it (Intersection) will give false results because 1:8 may not be a verse with a hit in 1:1-9, which reflects only the range of a clause with the hit(s).

MJ. Smith

https://community.logos.com/discussion/comment/438155#Comment_438155

which reflects only the range of a clause with the hit(s).

I'm not thinking specifically of passage lists that are claused based.

It is doable, but it (Intersection) will give false results

I agree it is not a perfect solution - the question is whether we can lower the number of false results. After doing some statistical analysis, one might well be able to create a better algorithm. However, one that produces fewer false positives without increasing false negatives AND is understandable to the average user is needed.

Dave Hooton

https://community.logos.com/discussion/comment/438165#Comment_438165

However, one that produces fewer false positives without increasing false negatives AND is understandable to the average user is needed.

That is the key.

I'm trying to manually edit the passage ranges from person:jesus across the whole NT and it is taking 2 hours + tedious checking when you get to Paul's convoluted statements!